Market housing creates affordable housing by displacing demand. No matter what kind of housing you build, as long as it gets built now it will lower prices now.

That assumes there are market incentives to actually fill the housing that is built. What about when buildings of hundreds of luxury condos are built and they just sit empty, either owned by disinterested "investors" or unwilling to be sold for the market-directed price?

Someone who studied this in Toronto by overlaying nighttime photos to find units where the lights are literally never on: https://www.movesmartly.com/articles/condo-units-sitting-emp...

That blogpost you cited says that there is a 5% vacancy rate. That's well within the normal range. So I don't see how it supports your argument.

https://www.reddit.com/r/badeconomics/comments/musne8/dispro...

Fair— I kind of picked one at random. Obviously part of the point of that dude's effort is that this is a difficult thing to get numbers on. Here's some other coverage around responses to a planned vacant unit tax:

https://www.blogto.com/real-estate-toronto/2021/02/real-esta...

Raise capital gains taxes on land. If they don't sell or rent it's their loss.

If you build more prices will come down and these "investments" will become nonviable.

Maybe, but that is difficult when a) the "investors" are split between connected, powerful people who really don't want to see their investments go negative; and families just getting started who could be utterly ruined by an underwater mortgage, and b) the fact that the housing situation is a crisis now; there isn't time to faff around and hope that ten to twenty years the invisible hand of the market does this or that.

I think you're mostly right, but it's not as straightforward as 1 dwelling per household and everyone just moves up a rung. Market housing could get bought up by people who are currently living with roommates (but would like their own home) or one of the 52% of young adults currently living with their parents.[0] Those spots are then unlikely to become available for the homeless.

This is just another symptom of the major lack of housing in the US, and needs to be solved as well, but it highlights how difficult it is to alleviate homelessness when there are other casualties of the housing market as well. By all means, build as much housing as soon as possible, but some of it will probably need to be specifically targeted at homeless or poorer populations.

[0] https://www.pewresearch.org/fact-tank/2020/09/04/a-majority-...

Why are those people living with roommates? Because they can't afford not to. If they suddenly move out and fill up a space they otherwise couldn't afford, that requires that prices have gone down.

price = F(supply, demand) therefore demand = G(price, supply) and supply = H(price, demand)

More luxery houses would lower prices for luxery housing, but not necessarily for affordable housing. I don't know how much those are competing markets

In addition to the relation between luxury housing market and affordable housing market that others are calling out, regulations and other hurdles to build are a reason expensive units are built rather than cheap ones. If you need 15 permits to build in SoMa and changes requested by one council can require going through the process for another permit again, your fixed cost gets so high that you really have no choice but to build expensive units. I'm sure housing cost would quickly get under control of SF for example were to eliminate any height and other zoning constraints and have as the only requirement that you are able to get insurance for your building.

New housing is always relatively "luxury" because newness is itself a luxury. Much of what is relatively affordable now was relatively "luxury" when it was built. But if you stop building anything new it would be like trying to keep the price of cars down by banning new cars – what will happen after a decade or two if you did that?

To use your car analogy- just because a new car will be worth more than a used one (and this is much more true of cars than houses), that doesn't mean there's no such thing as a luxery car. Sure, newness is a type of luxery (I'm not sure how much housing price depreciates with time, but I'll believe you on this one), but new houses can be built to cater to wealthy buyers or to cater to lower income buyers.

They compete strongly. Rich people pay a lot for junky apartments due to the shortage.

Somehow I doubt that's a considerable driver of affordable housing prices, but it's possible. Where did you get that from?

You ever seen techies living in the mission district?

Not what was said. A new building automatically makes the ten-year-old next door look shabby in comparison, and so on.

Still, we need to remove barriers that prevent middle-class housing from being built, such as parking minimums and other bureaucracy.

If you build enough luxury housing it will become cheap right away.

San Francisco has ~400k housing units. If you build a million more, prices will go down very sharply.

When there is more housing units than people who want to live there, prices go towards zero, whether they're "luxury" or not.

This is all true if we're talking about supply demand lines on a graph, but when on earth in real life is a for-profit company gonna whoops accidentally build so much that they have to take a loss and then keep building? lol.

I mean the idea is correct, but I don't think we can depend on the private market to deliver on this concept.

Don't get me wrong SF badly needs more housing and needs to approve tons more, I'm just saying there are limits to the outcome. At some point for profit developers will not build because they can't make a return.

Now if we had a not for profit government agency that steadily built housing regardless of the market conditions, say they were working toward a vacancy rate goal, now that could create surplus housing.

You might see friction there though in the private market would pull back.

I was arguing against the idea that adding "luxury" housing won't lower prices.

Most people understands it does it if you add a million units, but for some reason they can't grasp that each new unit contributes.

I was not predicting anyone would build a million units overnight, certainly not accidentally.

> At some point for profit developers will not build because they can't make a return.

SF has been very near this point for decades. Because new construction is mostly illegal, and if not extremely expensive and slow. Because of explicit policy from the SF BoS.

If housing construction were legalized in SF, say by adopting the regulations of Houston or Tokyo, there would be an enormous construction wave in the city. No one can predict to what level, but after that has transformed the city we can take stock and see if any other measures need to taken.

But I don't see SF doing anything of the kind without at least a decade of hesitation. The city is far too conservative to change.